Article Figures & Data
Figures
- Figure 1.
Association of IGF-IR levels with h10H5 response and ER status. A, 41 breast cancer cell line were screened for in vitro sensitivity to h10H5 using an ATP-based cell viability assay. Left axis and bar chart, IGF-IR mRNA level for each cell line as determined by gene expression microarray; right axis and diamonds, EC50 for h10H5 in each cell line; bottom, chart shows ER status for each cell line as determined by immunohistochemistry on a cell pellet tissue microarray. B, a combination of high expression of IGF-IR and the substrates IRS1 and IRS2 is associated with in vitro response to h10H5 in breast cancer cells. Heat map shows expression of IGF-IR, IGF-II, and the substrates IRS1 and IRS1 in breast cancer cell lines. Color coding is by z-scores. Red, high expression (2 SD above the mean); green, low expression (2 SD below mean); purple, cell lines that are sensitive to h10H5; yellow lines, cell lines that are insensitive to h10H5. C, pharmacodynamic response of sensitive MCF-7 and insensitive MDA-MB-231 cells to h10H5 treatment. Cells were treated with 1 μg/mL h10H5 for 24 h and lysates were used for immunoblotting with antibodies detecting the epitopes (right).
- Figure 2.
Diagnostic assays for patient stratification in clinical trials. A, agreement between protein staining intensity with an IGF-IR immunohistochemical assay with mRNA levels in 42 breast cancer cell lines. Points, cell lines; X axis, immunohistochemistry category (1+, 2+, and 3+); Y axis, IGF-IR mRNA levels. Examples of immunohistochemistry 1+ and 3+ staining are shown for the cell lines EVSA-T and BT474. B, examples of low (1+), moderate (2+), and high (3+) immunohistochemical staining in neoplastic breast tissue samples. C, distribution of low, moderate, and high immunohistochemical staining in a panel of breast and colorectal tumor samples. D, quantitative reverse transcription-PCR with a panel of biomarkers including IGF-IR, IGF-II, IRS1, and IRS2 was done on a set of FFPE colorectal tumors. The heat map is color coded by z-scores.
- Figure 3.
Combined effects of ER and IGF-IR targeting in vitro and in vivo. A, expression of IGF-IR and IGF-I in ER high and low human breast tumors and protein expression in ER-positive tumors. Heat map shows expression determined by Affymetrix microarray and is color coded by z-scores. B, effect of siRNA ablation of ESR1, the gene encoding ER, or IGF-IR siRNA ablation on mRNA levels of ESR1and IGF-IR in MCF-7 breast cancer cells. Cells were transfected with a control siRNA (NTC) or siRNAs targeting ESR1 or IGF-IR for 72 h, RNA was prepared, and IGF-IR levels were assessed by quantitative reverse transcription-PCR. IGF-IR is knocked down by IGF-IR siRNA treatment and also substantially reduced by ESR1 depletion. IGFBP2 is shown as a control to show that not all pathway components are down-regulated by ESR1 and IGF-IR treatment. C, effects of combined in vitro targeting of ER with the selective inhibitor fulvestrant and IGF-IR with h10H5. Cells were cultured in 2.5% FBS. Trastuzumab is included as an antibody control because MCF-7 cells are HER2-negative and do not show any response to anti-HER2-targeting agents. The combination of fulvestrant and h10H5 shows substantially greater inhibition of cell viability than either single agent. D, combined treatment with tamoxifen and h10H5 shows superior tumor growth inhibition to either single agent in xenografted MCF-7 tumors. Exogenous estrogen was provided in drinking water. Arrowheads, h10H5 was administered weekly; arrow, tamoxifen slow-release pellet was implanted at the start of the study.
- Figure 4.
Association of IGF-IR levels with in vitro h10H5 response in colon cancer. A, 27 colorectal cancer cells line were screened for in vitro sensitivity to h10H5 using an ATP-based cell viability assay. Left axis and bar chart, IGF-IR mRNA expression levels determined by microarray; right axis and diamonds, EC50 for h10H5 in each cell line. B, percent inhibition of in vitro cell viability by h10H5 (X axis) is correlated with IGF-IR mRNA levels determined by microarray (Y axis). Points, cell lines. C, pharmacodynamic response of sensitive HT-29 and insensitive HCT-116 cells to h10H5 treatment. Cells were treated with 1 μg/mL h10H5 for 24h and lysates were used for immunoblotting with antibodies detecting the analytes (right).
- Figure 5.
A gene expression signature of biomarkers of response to h10H5 in colorectal cancer cell lines. A, heat map showing expression of 60 genes identified through supervised analysis of gene expression data that distinguish h10H5-sensitive colorectal cells from resistant cells. Y axis, genes; data were derived from log transformation and median centering for each gene. Red, high expression; green, low expression according to z-scores. B, relationship of expression of a single candidate predictive biomarker, CD24, with growth-inhibitory effects of h10H5 in cell lines. Blue columns, CD24 mRNA expression; red diamonds, percent inhibition of cell viability observed in response to 1 mg/mL h10H5 treatment over 3 d. Bars, SD from four replicate experiments. C, schematic of various classes of genes implicated in the h10H5 sensitivity and proposed relationship to signaling through the IGF-IR axis.
- Figure 6.
Activity of h10H5 in colorectal xenograft and primary tumor explant models. A, Colo-205 tumors cells and CXH-280 primary colorectal tumor explant tissue were profiled on gene expression microarrays and data are shown for IGF-IR and the IGF-II. Colo-205 is a high receptor expression model and CXF-280 is a high ligand-expressing model. B, 14-day daily dosing of flank xenografted Colo-205 high IGF-IR cells with h10H5 substantially reduced tumor growth in a dose-dependent manner. C, 14-day daily dosing of the human primary tumor explant xenograft model CXF-280 with h10H5 resulted in substantial reduction of tumor growth compared with animals dosed with vehicle or a control antibody.
Additional Files
Commentary on this Article
The following Commentary is available for this article:
Predictive biomarkers for targeting insulin-like growth factor-I (IGF-I) receptor
Craig P. Carden, L. Rhoda Molife, and Johann S. de BonoSupplementary Data
Files in this Data Supplement:
Volume 8, Issue 8
